[cinwendy]

Hi,

I am trying to use some commands to save contour file for transient stimulation. I saw a lot of examples here. But I couldn't found what I should do if I don't know the max and min value beforehand. I do not want a too big or too small max/min value and get a "distorted" contour plot.

What I do using TUI:
/display> contour
contours of> uds-0-scalar
minimum value [645.43408203125] (press enter)
maximum value [1184.197631835938] (press another enter)

While the way of doing this in command is typically
/display/contour mixture velocity-magnitude 0 0.5
that means the min and max values are known.

How to put the "enter" into command? Empty spaces, "yes", empty brackets all doesn't work.

Thank you in advance!

[blackmask]

What about two empty new lines?

The thing is if the min/max for the contours changes during the simulation, how do you compare your results? For example, the color red in the first image represents a value to 1200 when the range is automatically determined as [1000,1200], but the same color in the second image could represents a value of 1000 where the range is automatically determined as [1000,1100].

I suggest that you estimate the proper value for the max/min and set them beforehand even if the range could be dynamically determined.

[cinwendy]

I see. Ok, I will try some engineering guessing.

But... another issue is bothering me. As I was looking through my other contours, I realised that for some of them, not all the colours are shown. I checked on every surfaces but all that are showing up are either just the red-end or the blue-end colours. The value at the other end that was not showing up is the one that bothers me. To my understanding of my problem, those values are very undesirable, some of them are simply unrealistic. This is also the reason I started this thread-- these values make me feel hard to guess beforehand.

Does these values really exist somewhere? How can I find them? Actually I kept getting "divergence detected in AMG solver" error, and still trying to get rid of it. These contours are the only one I get without any error, somehow after a lot of clicking somewhere in FLUENT. So, it is kind of hard to guess a proper range. And, I am also trying to check if the results really make sense.

So, how can I find those missing colours? Can I ignore it? How to judge?

Thank you!

[blackmask]

Search the phrase "divergence detected in AMG solver" in this forum and you will find a lot of threads discussing why these messages are reported and how to avoid them. The most common situation is that you have a low mesh quality somewhere in your computational domain. You should deal with this problem first because otherwise the result might be not meaningful.

[cinwendy]

I just get rid of the "divergence in AMG solver" error by properly initializing my simulation, the only way that worked so far. Now that I get some "converged" results, so I am looking at those strange contours...

[blackmask]

Do not use the global range or auto range. Use the clip range option and adjust the range until what you feel appropriate.

[diamondx]

try (), it works sometimes

[niloogh]

Originally Posted by cinwendy

I just get rid of the "divergence in AMG solver" error by properly initializing my simulation, the only way that worked so far. Now that I get some "converged" results, so I am looking at those strange contours...

hi
i have the same problem as you:Error: divergence detected in AMG solver
would you plz tell me how did you solve your problem????